In the debate around Governor Brown’s proposed “Local Control Funding Formula” (LCFF), a number of issues have been raised that school finance researchers (and policymakers in other states) have been discussing for years. This ‘School Finance 101’ series will highlight what we know – and what we don’t know – about some of these issues. Much of what is covered in these posts (including additional sources and full citations) can be found in School Funding Formulas: What Works and What Doesn’t? Lessons for California, a 2007 report done for the Senate Office of Research.
Across the country, the primary goal of most state finance systems is to promote equalization, particularly in states where locally-financed school systems have faced court challenges. Certain formulas achieve this goal better than others. As noted in my post yesterday
, both California’s current revenue limit system and Governor Brown’s proposed formula are versions of a traditional foundation
state-aid formula. Typically, in a foundation system, the state assumes (or requires) that each district levies a minimum tax rate. If local revenue raised at that rate is less than the foundation amount, then state aid makes up the difference. Thus, basic state aid is defined as:
State aid per pupil = Foundation amount per pupil – (required tax rate*assessed property wealth per pupil)
In our current system, a district’s revenue limit acts as the foundation amount and the ‘required tax rate’ is determined by the state formulas for allocation of property tax revenue. It should be noted that although Prop 13 sets the total tax rate on property across California at one percent, the effective rate for school districts is less than that because revenue is shared with other local governments. Due to the complexities of the property tax revenue allocation formula, the effective tax rates for districts range from 0.1 to 0.6 (Sonstelie, 2001
What process determines the foundation amount?
The majority of states now use foundation formulas, either solely or in combination with other systems. But in most states, equalization of foundation amounts has been accomplished by spending more in the poorest school districts, i.e., ‘leveling up’, and thereby increasing total spending on education. In this regard, California is an exception: although per-pupil spending was substantially equalized after Serrano, average spending in California actually fell and today, California consistently ranks toward the bottom in per-pupil spending. It is therefore not surprising that one of the big questions about Governor Brown’s proposal is what the base amount should be. Conceptually, there are a number of ways that policymakers could decide on a foundation level, with the three most relevant being History, Politics and Cost.
History: Currently, the base level of unrestricted revenue in most California districts is determined almost entirely by the revenue limit. The amounts themselves are generally based on history; i.e., expenditure levels in the 1970’s, with adjustments for inflation and equalization. There are some adjustments for necessary small schools and district type (unified, elementary, high school) but for the most part, the amount allocated to a district under the current system need not bear any connection to the characteristics of the district or to the amount actually required to meet the costs of achieving any desired performance goal.
Politics: Governor Brown has proposed base rates for his funding formula that range from $6432 to $7680 depending on grade level. These values were based on the state average revenue limit amounts, after correcting for budget-induced reductions over the last several years. One way to interpret this is that these base values are a result of a political calculation (i.e., what makes political sense) and budgetary reality (i.e., what we can afford). To be fair, this is also how the base foundation amount has been determined in most other states.
Cost: In theory, the foundation amount could be set according to some conceptual model of the educational needs that schools face. That is, rather than the base revenue per pupil in each district being determined by some ad hoc historical or political guideline, the foundation spending per pupil could be derived from the actual cost for a district to achieve the performance outcomes the state has established. This is often referred to as an “adequacy” approach, the idea being that the base is set to be “adequate” for providing some minimum level of educational quality. For example, in 2002, Maryland adopted significant school finance reforms that relied heavily on a ‘costing-out’ study completed by an independent consulting firm and commissioned by a bi-partisan Commission. Wyoming, Arkansas, and New York have also set the foundation amounts in their formulas based on cost studies performed by independent consultants, while Oregon, Washington and Maine have all developed cost models internally (i.e., by state commissions rather than external researchers). In most cases, these models are re-calibrated periodically to ensure that they continue to reflect state standards, goals and cost variations. In addition, a large number of states have commissioned cost studies, or cost studies have been completed as part of adequacy litigation, even if the findings have not yet been directly incorporated into funding levels.
What should the base be?
There can be little doubt that if
one is concerned about efficiently maximizing student outcomes (or avoiding an adequacy lawsuit), then it is preferable to set the minimum foundation level based on the actual costs of educating students. What is less straightforward is how to determine what those costs are. There is a growing literature on estimating the costs of education; a full discussion is well beyond the scope of this post but most studies involve one of four methodological approaches: professional judgment, evidence based, successful schools, or the cost function (or “econometric”) approach. See Duncombe and Yinger (2004)
and Baker, Taylor and Vedlitz (2004)
for a full discussion of the advantages and disadvantages of each of these methods.
The Getting Down to Facts project
included three ‘costing-out’ studies for California: Chambers, et al (2007)
used a traditional Professional Judgment approach in which two panels of educators were convened and asked to construct a prototypical school that would meet state standards; Sonstelie (2007)
used a modified Professional Judgment approach in which 567 teachers and principals were surveyed about their optimal configuration of resources; and Imazeki (2007)
estimated an econometric cost function which uses data from all districts in the state to identify relationships between district characteristics, spending and outcomes.
The three GDTF cost studies produced varying estimates of the minimum cost for a basic district (i.e., with low levels of student needs) to meet the state standard of an Academic Performance Index of 800; the estimates ranged from slightly under $6000 to around $8400 per pupil in 2004-05, or $7122 to $9971 in 2011-12 (using the CPI-U to adjust for inflation). Imazeki also included a summary of base costs in a number of costing out studies done in other states. Although it is difficult to compare these estimates across studies, due to differences in methodologies, state standards and populations, those estimates average between $7371 and $9365 (in 2011-12 dollars).
LCFF proposed base too low?
Given these numbers, one could certainly argue that the base amounts in Brown’s proposal are a bit lower than necessary to cover the minimum costs of educating California students. That argument is further bolstered when one considers that, regardless of methodology, cost studies use past data and experience to extrapolate what districts will need in the future to reach some specified goal. If the educational environment (including the accountability and assessment institutions) change, then those estimates will be less helpful. Considering that California schools are headed into uncertain territory with the switch to the Common Core and new accountability measures, it is likely that past cost estimates are more of a lower bound on the true cost of achieving future performance goals.
LCFF proposed base sufficient?
On the other hand, Brown’s proposed base amounts are not significantly lower, and in some cases they are much higher, than the foundation amounts in other states. A 2009 survey
shows that the base levels in other states that use a foundation formula ranged from a low of $1721 ($1807 in 2011-12 dollars) in Oklahoma to a high of $9678 ($10,162 in 2011) in Connecticut, but most states fall between $3000 and $6000 ($3150 and $6300 in 2011). New York’s foundation amount was $5695 ($5980 in 2011); Florida and Texas had even lower bases ($4171 and $3292, respectively, in 2011) but those states also have formulas where the base is adjusted upward for a larger number of district and student characteristics.
It is important to keep in mind that the base amount in a foundation formula is NOT the same as the average per-pupil revenue – no district will receive the base amount per pupil and NO other revenue. Base funding levels should not be compared to ‘average spending’ in other states, which almost certainly include additional revenue for additional costs. In a cost framework, the base amount represents the minimum cost to educate students in a district with NO students with special needs and no other characteristics (like size or location) that would raise costs relative to other districts. Virtually all states have adjustments for some student and/or district characteristics, which I will discuss in the next few posts.